Targeting
One agreed list of companies that fit, scored on your own won and lost deals and tiered for ABM and ABS.
ABM and ABS setupIdeal customer profile playbookIdeal customer finder
Playbook · GTM engineering
GTM engineer: the role and the first 90 days
What a go-to-market engineer builds, the skills it takes, and a week-by-week plan that shows results by day 90.
- For
- New GTM engineers, RevOps and founders hiring one
- Time
- 12 weeks
- You get
- A running engine and a list of what to automate next
What is a GTM engineer?
A GTM engineer builds the systems behind sales and marketing: who to target, when to reach out, how email lands, how leads are routed, which AI agents do the prep and how results are measured. Clay’s head of GTM engineering, Everett Berry, describes the job as removing the process and technology blockers between a go-to-market idea and running it at scale.
Commercial
Knows the buyer
Who fits, why now, what to say.
Technical
Builds the system
Data, workflows, APIs, agents.
Operational
Keeps it running
Clean data, approvals, measurement.
What a GTM engineer builds
Signals
Timing: who changed, who is researching, who came back, routed to the right owner.
Outbound infrastructure
Domains, mailboxes, authentication, warm-up and sending limits that keep email in the inbox.
AI agents with approval
Research, prep and follow-up done by agents; people approve what is sent or saved.
CRM and routing
Clean data, clear lifecycle stages, fast hand-offs and no lead without an owner.
Measurement
Which channels and plays turn into meetings and revenue, not just activity.
How a GTM engineer sets up ABM and ABS
Account-based marketing (ABM) aims marketing at a named list of companies. Account-based selling (ABS) is the sales side: researching each account and reaching several people in its buying group at the right moment. The GTM engineer builds the one system both teams run on: the same account list, tiers, signals and CRM view.
| Tier | Which accounts | Research | How they are worked |
|---|---|---|---|
| Tier 1 · one to one | The few dozen best-fit accounts | Full account brief and buying-group map | A plan per account, personal outreach to several people, ads and content for that account |
| Tier 2 · one to few | Clusters by industry or trigger | Short brief, one shared angle per cluster | A sequence per role, ads per cluster, events |
| Tier 3 · one to many | Every other company that fits | Fit and signal score only | Signal-triggered sequences and retargeting |
- Build the account list.Score fit on your own won and lost deals, add "not a fit" rules, and agree the list with sales and marketing.
Ideal customer finder skill → - Tier it.Tier 1 is the few dozen accounts sales can work by hand. Review the tiers every quarter, not every week.
Lead scoring designer skill → - Map the buying group.For tier 1 and tier 2: roles, not just titles. Who decides, pays, uses and can block.
Stakeholder mapping playbook → - Score signals per account.Add up fit and timing at the account level, not per lead, and let older signals count less.
Buying signal monitor skill → - Brief before outreach.One page per account: what changed, why now, and the angle for each person.
Account brief skill → - Run plays together.Marketing warms the account with ads and content to the list; sales reaches three or more people with a message per role, timed on signals.
Cold email writer skill → - Put it in the CRM.Tier, score, owner and next step on the company record. One owner per account, one view for both teams.
CRM hygiene audit skill → - Measure per account.Engaged accounts, meetings, pipeline and win rate per tier. Not leads per form.
Channel attribution skill →
- Account listFit on won deals
- Tiers1:1, 1:few, 1:many
- Buying groupRoles per account
- SignalsScored per account
- ABMAds and content to the list
- ABSSeveral people, a message per role
Skills
| Area | What good looks like |
|---|---|
| Commercial judgement | Can explain the ideal customer, the buying group and why a signal means "now". |
| Data | Spreadsheets, SQL basics, APIs and webhooks; knows where each field comes from. |
| Workflow tools | Clay, n8n or Zapier; builds once, documents, reuses. |
| CRM | Properties, lifecycle stages, routing and reports in HubSpot or Salesforce. |
| Deliverability | Domains, SPF, DKIM, DMARC, warm-up and sending limits. |
| AI agents | Writes skills, connects tools through MCP, sets approval steps and guardrails. |
GTM engineer vs RevOps
RevOps
Runs and measures the revenue process
Stages, forecasting, reporting, tool policy, compensation data.
GTM engineer
Builds the pipeline machine
Data, signals, outbound infrastructure, automation and agents.
In startups and small teams one person often does both. The split matters once each needs full-time attention.
The first 90 days
Build in this order. Each step makes the next one cheaper, and each week ends with something that works, not a slide.
Days 1 to 30 · Foundation
- Week 1List every tool, data source and owner. Buy sending domains on day one: warm-up takes 3 to 5 weeks.
Inbox calculator → - Week 2Audit the CRM: duplicates, missing owners, stale deals. Fix what breaks decisions first.
CRM hygiene audit skill → - Week 3Compare won and lost deals to define the ideal customer profile, with "not a fit" rules.
Ideal customer finder skill → - Week 4Authenticate every sending domain and agree suppression rules: customers, open deals, competitors.
Email deliverability test →
Days 31 to 60 · Engine
- Week 5Set up ABM and ABS: tier the account list, then map the buying group for every tier 1 account.
ABM and ABS setup → - Week 6Pick three signals you can see reliably and route each to an owner.
112 buying signals → - Week 7Build a lead score from your own history and test it on past deals before launch.
Lead scoring designer skill → - Week 8Launch one campaign: one segment, one angle, three steps. For tier 1, run it as an ABS play: several people per account, a message per role.
Cold email writer skill →
Days 61 to 90 · Scale
- Week 9Make sure positive replies and inbound leads reach a person within the hour.
Lead hand-off check skill → - Week 10Add agents for research, call prep and follow-up, with approval before anything is sent or saved.
Free Claude skills → - Week 11Report meetings and pipeline by segment, signal and channel.
Channel attribution skill → - Week 12Write down every process you ran twice, as a skill or a recipe someone else can run.
How our skills are built →
- AuditTools, data, CRM
- TargetICP, tiers, ABM
- SignalsTiming and routing
- OutboundOne campaign
- AgentsWith approval
How to measure it
Measure the system, not just the volume. Five numbers, reviewed weekly: meetings booked by segment and signal; time from lead to first human response; share of sends that bounce; share of CRM records with an owner and a valid stage; and pipeline created by play.
What goes wrong
- Tools before targeting. Automating outreach to the wrong companies only fails faster.
- A month of meetings and no build. Ship one working thing every week.
- Fourteen tools stitched together. Fewer tools, one source of truth, documented flows.
- ABM as an ad campaign. Ads to a list without sales working the same accounts at the same time is only awareness.
- AI on dirty data. Agents repeat whatever the CRM says. Clean first.
- Counting activity. Emails sent is not a result; meetings and pipeline are.
Questions people ask
What is ABM, and how is it different from ABS?
Account-based marketing (ABM) aims marketing at a named list of companies instead of anyone who fills in a form. Account-based selling (ABS) is the sales side: researching each account and reaching several people in its buying group at the right moment. They work best as one system, with the same account list, tiers and signals for both teams.
Do you need to code to be a GTM engineer?
Not to start. Most of the work happens in tools such as Clay, n8n, Zapier and the CRM. Some SQL, scripting and comfort with APIs helps once workflows get bigger, and AI assistants like Claude Code make that part much easier to learn on the job.
What does a GTM engineer do day to day?
Builds and maintains workflows that find, enrich and score accounts, keeps outbound and signal-based campaigns running, protects data quality in the CRM, and runs experiments to improve reply and conversion rates. A growing share of the work is setting up AI agents with clear approval steps.
What is the difference between a GTM engineer and RevOps?
RevOps owns how the revenue process runs and is measured: stages, forecasting, reporting, tools and policy. A GTM engineer builds the systems that create and act on pipeline: data, signals, outbound and automation. In small teams one person often does both.
What skills does a GTM engineer need?
Commercial judgement first: who to target, why now and what to say. Then data skills (spreadsheets, SQL basics, APIs and webhooks), workflow tools such as Clay or n8n, CRM administration, email deliverability, and the ability to build and govern AI agents.
When should a company hire a GTM engineer?
When the team knows who to sell to but cannot reach them at the right moment at scale: manual list building, slow follow-up, signals that nobody acts on, and tools that do not talk to each other. Before that, fix positioning and the ideal customer profile.